Our long-term goals are to find effective preventive and therapeutic approaches for a group of carcinomas that are disproportionately responsible for cancer mortality worldwide. This proposal will seek to identify the cell of origin for a group f aggressive malignancies from disparate sites that share a common etiology of a metaplastic response to chronic inflammation. Our previous studies, and the novel insights they have generated, provide us with a compelling hypothesis that defies the current dogma and classification of these selected malignancies. We propose that select subsets of carcinomas, exemplified in this proposal by esophageal adenocarcinoma, gastric carcinoma of the indeterminate type and metaplastic breast cancers, arise from unique phenotypically plastic cells that confer aggressive phenotypes with extremely poor prognosis. These subsets of carcinomas are typified by the presence of metaplastic tissue derivatives in the pre-malignant and malignant lesions. Metaplasia, a protective adaptive tissue response where one cell type is replaced by another cell type not typically present in that tissue, is indicative of chronic tissue injury and associated with increased incidence of cancer. Unraveling the biology of these metaplastic carcinomas will allow us to overcome barriers to the identification of predisposing conditions and the subsequent activation of the cells-of-origin leading to lethal metaplastic carcinomas and to develop biomarkers to identify those premalignant lesions that will progress to invasive cancer. Identification and targeting of signaling pathways that control progression of these cells would allow us to design rational therapeutics that target the invasive state of these carcinomas. Finally, we also propose a novel preventive approach to reduce formation of the pre-malignant state that underlies these malignancies. Realization of our goals would change clinical practice and allow physicians to halt these cancers before they form. We propose to: (1) determine the role of specific stem/regenerative cells in initiation and progression esophageal adenocarcinomas, gastric cancers of indeterminate type and a subset of triple negative breast cancers, i.e. metaplastic breast cancers. We will also (2) identify distinct molecular features (cellular states) and match them with an optimal approach for (3) targeted prevention and (4) therapy. The connection between chronic inflammation and malignancy is well known; however despite an understanding of the causal insults, the incidences of these malignancies are increasing in the modern world and significantly contribute to cancer mortality. If successful, the goals of this proposal would delineate a therapeutic paradigm that could reduce mortality from this subset of cancers and fundamentally alter our approach to cancer prevention.